code-review-graph

Query codebase structure graphs for impact analysis and semantic search via CLI or mcporter.

2|3|Updated Mar 13, 2026
One-click install
npx skills add https://github.com/spideynolove/claude-dotfiles --skill code-review-graph-spideynolove
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: code-review-graph
Source: https://github.com/spideynolove/claude-dotfiles/tree/main/.agents-global/skills/code-review-graph
Command: npx skills add https://github.com/spideynolove/claude-dotfiles --skill code-review-graph-spideynolove

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Reviewing code changes without structural context leads to missed dependencies and broken callers. This Skill gives you graph-backed codebase maps, impact analysis, and semantic search without keeping the MCP server natively registered in Codex. ## Core Features & Use Cases - Graph Status & Builds: Run code-review-graph status and code-review-graph build to check or rebuild the local codebase graph. - Change Detection: Use code-review-graph detect-changes --brief and update --skip-flows to keep the graph fresh after edits. - On-Demand MCP Tools: Call specific MCP tools through mcporter (npx mcporter call code-review-graph.TOOL_NAME key=value) without registering the server natively. - Use Case: Before modifying a shared utility function, run impact analysis through the graph to see every caller and dependent module that your change could affect. ## Quick Start Ask the AI to run code-review-graph status and then perform an impact analysis on the function you are about to modify.

Frequently Asked Questions about code-review-graph

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I check the status of my codebase graph?

Run `code-review-graph status` from the CLI to check the current state of the local graph. Use `code-review-graph build` if the graph needs to be created or rebuilt from scratch.

How to use code-review-graph MCP tools without registering the server?

Use mcporter to call tools on demand: run `npx mcporter list code-review-graph --schema` to see available tools, then `npx mcporter call code-review-graph.TOOL_NAME key=value` to invoke one. This avoids keeping the server natively registered in Codex.

When should I use the CLI versus mcporter for code-review-graph?

Prefer the CLI for routine work like status checks, builds, updates, and change detection. Use mcporter only when a specific MCP tool is needed, since it invokes tools without exposing the server as a native Codex MCP.

Does code-review-graph update automatically after code edits?

Yes, the global Codex hooks run `code-review-graph build` or `status` on session start and `code-review-graph update --skip-flows` after Bash or apply_patch edits, keeping the graph fresh without manual intervention.

What is the difference between update and detect-changes in code-review-graph?

`code-review-graph update --skip-flows` refreshes the graph after edits while skipping flow analysis, making it faster. `code-review-graph detect-changes --brief` reports what changed in the codebase in a compact format.